mGlu ligands for treating PD: mGluR5 antagonist & mGluR4 PAM
Parkinson Disease has remained to be one of the biggest issue of the world mainly due its high prevalences. Parkinson Diseases was found to be caused by dopamine depletion, which causes an imbalance in the glutamate and GABA synapses. It was thought that dopamine depletion lead to hyperactive GABAergic synaptic transmissions, which increases glutamatergic firing in the SNr that ultimately causes motor symptoms. Current Parkinson treatment uses dopamine precursor L-DOPA to relieve symptoms. However, long-term L-DOPA treatment induces dyskinesia. In this review, mGluR 5 antagonist and mGluR 4 positive allosteric agonist are evaluated by their promising potential to relieve or even treat Parkinson diseases.
Parkinson Disease (PD) has remained to be one of the biggest worldwide issue. Each year, 60,000 Americans are diagnosed with PD, and more than 10 million people worldwide currently have PD1. PD was found to be caused by dopamine depletion, which causes an imbalance in the glutamate and GABA synapses2. It was thought that dopamine depletion lead to hyperactive GABAergic synaptic transmissions, which increases glutamatergic firing in the SNr. The net result is increase in glutamate level in the SNr region then ultimately lead to PD motor deficits2.
If you need assistance with writing your essay, our professional essay writing service is here to help!
Current treatment of PD uses dopamine agonist and dopamine precursor L-DOPA, but long-term L-DOPA treatment could lead to L-DOPA induced dyskinesia (LID), in which a patient experiences abnormal involuntary movement (AIM). Thus, many researchers sought to look for a non-dopaminergic pathway that could treat PD without inducing L-DOPA side effects. Researchers sought to approach this issue by studying the metabotropic glutamate receptors (mGluR) because they were found to mediate GABA and glutamate transmissions. Specially, heavy emphasis was put on group III mGluR because it is found in pre-synaptic terminal and become hyperactive when dopamine depletes2. It was also found that two main mGluR targets could potentially treat PD: mGlu5 anatomist and mGlu4 positive allosteric agonist (PAM). By literature, mGlu5 was found to reduce the high nigral GABA level that was observed during dyskinetic abnormal involuntary movements (AIM) and mGlu R 4 could inhibit the overactive glutamatergic and GABAergic synapses, restoring the synaptic balance2,8.
Classification of mGluRs AND detailed emphasis on mGlu4 and mGlu5 targets
mGluRs are G-protein-coupled receptors for glutamate. There are 8 subtypes of mGluR in the brain, which are classified into 3 groups by DNA sequence similarity and ligand specificity 2. Group I include mGluR1 and mGluR5, Group II include mGluR2 and mGluR3, and Group III include mGluRs 4, 6, 7, and 8.
Group I mGluRs mainly affect the postsynaptic transmissions and are coupled to phospholipase C, which cleaves PIP2 into DAG and IP3. The freely diffusible IP3 would activate cytosolic Ca2+ influx, and ultimately activates protein kinase C4. Group I mGluRs are also associated with the sodium and potassium channels, and their activity could lead to either (1) an excitatory signal to cause presynaptic neuron to release glutamate or (2) inhibitory signal to inhibit postsynaptic release of glutamate.
The public has put high attention on mGluR5 when it comes to alleviating PD symptoms. This is mainly because mGluR5 was found to mediate dopamine and glutamate release at central synapses. Moreover, a study that investigated mGluR5 found that selective mGluR5 antagonist (co-administered with L-DOPA) significantly alleviate the L-DOPA induced dyskinesia (LID) AIM symptoms. Mechanistically, this could be because mGluR5 antagonist reduces the increased nigral GABA level that was found in L-DOPA induced AIMs8.
On the other hand, Group II and III mGluRs mainly affect the presynaptic transmissions and are coupled to the inhibitory G protein, which inhibits adenyl cyclase, reducing the cAMP levels 3. Group II and III are also found to mediate the glutamatergic and GABAergic synapse transmission in the basal ganglia, and stimulation of group III mGluR was found to reduce glutamate and GABA release. Agonist of Group II and III would lead to the downstream inhibition response of adenyl cyclase, which reduces cAMP production. mGlu 4, 7, and 8 are found on the pre-synaptic terminals of basal ganglia pathway, which was found to be overactive during PD motor symptoms and dopamine depletion. Activation of group III mGlu R was found to inhibit the synaptic transmission (glutamatergic and GABAergic neurotransmission) that were found to be overactive during dopamine depletion. Fixing the GABAergic and glutamatergic balance in the basal ganglia could be a potential treatment to relieve PD symptom or prevent PD occurrence, and group III mGluR seem to be a potential target for fixing this balance because they are found on pre-synpatic terminals of GABAergic and glutamatergic neurons, making group III mGlu R a potential target to investigate to treat PD symptoms2.
Heavy focus was put on mGluR4 as a potential target for PD symptoms because agonist of activated mGluR 4 reduces GABA releases (which was found to be hyperactive during dopamine depletion). Another explanation could be because the other two group III mGluRs (mGluR7 and mGluR8) found in the hyperactive pre-synaptic terminals during PD symptoms weren’t fully characterized and no highly selective ligands for these subtypes were discovered2.
A mGluR5 PAM that reduce LID incidents in PD rat model
mGlu5R antagonist and L-DOPA co-treatment reduces AIMs and LID incidences
A group of researchers conducted further investigation on mGluR 5 effect on PD rats and discovered that antagonist of mGluR 5 reduces the dyskinesia induced by long-term L-DOPA usage in rat PD model. In this study, mGluR 5 antagonist was tested on 6-OHDA-lesioned rats, a common animal model for PD. Rats were injected with L-DOPA co-administered with 3-[(2-methyl-1,3-thiazol-4-yl)ethynyl]pyridine (MTEP). Results showed that treatment by L-DOPA alone induces AIMs, whereas co-treatment of L-DOPA and MTEP significantly reduces AIM symptoms. Chronic co-treatments show a reduction of AIMs by more than 60% than for the PD models treated by L-DOPA alone8. Through microdialysis, they found that AIMs correlate with increased GABA level in the substantia nigra pars reticulata (SNr). And co-administration of MTEP with L-DOPA reduces the nigral GABA levels. This suggest a possibility that the reduction of nigral GABA level by co-treatment was what actually suppresses the AIM symptoms 8.
This study also tested if the co-treatment physically suppresses dyskinesia symptom or if it actually prevents dyskinesia incidents. This was tested by washing out the drug effects and then injecting mice with acute L-DOPA and testing the dyskinesia symptoms. Results show that mice with no L-DOPA and MTEP co-treatment or treated by only MTEP have AIM scores significantly lower than mice with chronic L-DOPA treatment only (before the drug washout). As for the co-treated L-DOPA and MTEP mice, they had an 87% AIM score reduction when compared with the mice previously treated with L-DOPA. A hypothesis came up suggesting that dyskinesia level depends on GABA release in striatal efferent projection. So, the SNr GABA levels was measured. For L-DOPA injected mice, an increase in GABA levels in the SNr was observed along with the gradual increase dyskinesia AIM symptoms. When co-treated with MTEP, the GABA level in the SNr significantly reduced, along with the reduction of dyskinesia AIM symptoms, suggesting that MTEP relieve LID AIM symptoms by inhibiting the high GABA release (induced by L-DOPA) in the SNr. Overall, MTEP alone was found to not produce anti-akinetic effects in PD models, but its co-treatment with L-DOPA could prevent LID development, as MTEP inhibit GABA level and doesn’t interfere the therapeutic function of L-DOPA8.
The high potential behind Reason for Further Pursue on mGlu5R Antagonist
This study also found that the high nigral GABA level is what ultimately lead to LID in PD . This was found by comparing the nigral GABA level for dyskinetic and non-dyskinetic mice after treatment of: L-DOPA only, mGlu5R antagonist MTEP, the co-treatment of both, and vehicle group on PD mice models. Results show that only LID mice has high nigral GABA level after L-DOPA treatment [Fig 1a]. This was further confirmed by looking in the non-LID groups, in which L-DOPA only treatment didn’t lead to increase in nigral GABA level, suggesting that L-DOPA treatment lead to high nigral GABA level, which potentially aids LID development [Fig 1b]. The scientific explanation was that high GABA release might cause a reduction in neural activity associated with a change in firing pattern in basal ganglia output, which might contribute to LID. This discovery drives more researchers to further investigate mGLuR5 antagonist, which leads reduction in nigral GABA level, as treatment for PD8.
dissertations.com/108/0500817.001.png" alt="" />dissertations.com/108/0500817.002.png" alt="" />
[Figure 1a: SNr GABA level for 6-OHDA-lesioned rats that developed L-DOPA induced Dyskinesia (LID) after treatment(s) of: vehicle, L-DOPA only, MTEP, or L-DOPA + MTEP on 6
Figure 2a: SNr GABA level for 6-OHDA-lesioned rats that did not developed L-DOPA induced Dyskinesia (LID) after treatment(s) of: vehicle, L-DOPA only, MTEP, or L-DOPA + MTEP on 6] 8
mGluR4 PAMs that reduce LID incidents in PD rat model & How mGluR4 PAMs progresses from (-)-PHCCC to PXT002331
The rising potential of mGluR4 PAM in treating PD
While some researchers study in mGlu5R’s potential in treating PD, other researchers investigate on mGlu4R’s potential regarding PD. A group of researchers approaches the PD problem through targeting group III mGluRs because it mediates the glutamatergic and GABAergic neurotransmissions that were found to be hyperactive during dopamine depletion in PD2.
Activation of mGlu4R inhibit striatopallidal synapse transmission
One of the key experiment sought to study the effect of mGluR III agonist on the striatopallidal IPSCs (because increased striatopallidal IPSCs was thought to cause motor symptoms in PD). More specially, mice were treated with selective group III mGluR agonist L-AP4 and the striatopallidal IPSCs were measured. Results show that activation of mGluR III inhibits striatopallidal IPSCs. This was further confirmed when the drug was washed out, in which striatopallidal IPSCs amplitude went back to normal (before the drug treatment) 10. This result raises another question: which group III mGluR subtype is responsible for this effect?
Through series of experiments, it was discovered that mGluR4, which was found in the pre-synapse of striatopallidal terminals, could potentially be the group III mGluR that led to striatopallidal IPSC inhibition. This was suggested by compare the effect of L-AP4 on mGluR4 knock out (KO) mice and the vehicle. Results show that L-AP4 in control mice did lower striatopallidal IPSC. As for the mGluR4 KO mice, L-AP4 didn’t significantly reduce the striatopallidal IPSC, suggesting that mGluR4 activation inhibit the striatopallidal synaptic transmission 10.
Recall that the increased GABAergic inhibition at the inhibitory striatopallidal synapse was thought to cause motor symptoms. Since activating mGluR4 could inhibit the hyper striatopallidal IPSCs that was observed in PD motor symptoms, this made mGluR4 agonist a high potential candidate to look into when researching for PD treatment10.
A mGlu4R PAM in mice models
From (-)-PHCCC to compound 40:
After the study regarding striatopallidal IPSC inhibition, many researchers start looking into mGluR4 agonists for PD research. However, selective targeting on mGluR4 was ‘historically’ proved to be difficult. In the past decade, mGluR4 has already been a high potential target for treating PD mainly because activation of mGlu4 was found to reduce they hyperactive GABA and glutamate release in pallidal segments and inhibit the glutamatergic and GABAergic neurotransmissions that were found to be hyperactive during PD motor symptoms. However, there hasn’t been mGluR ligand that has been proven worthy to enter into the clinical trial except the recent mGlu4 PAM PXT002331 9.
Our academic experts are ready and waiting to assist with any writing project you may have. From simple essay plans, through to full dissertations, you can guarantee we have a service perfectly matched to your needs.
One explanation for such delay was the difficulty to develop a high mGlu4 specificity and brain penetrant mGlu4 ligand. (-)-PHCCC ((−)-N-Phenyl-7-(hydroxyimino)cyclopropa[b]chromen-1a-carboxamide) was one of the first mGluR4 PAM described in literature. However, (-)-PHCCC is not a good drug candidate because of its low selectivity for mGluR4 and poor penetration across the blood brain barrier. Through a series of modifications of this compound, a group of researchers successfully discovered compound 40 (PXT001687), which was tested to have higher brain penetration and selectivity for the mGluR4 PAM than the (-)-PHCCC compound9.
Compound 40 has L-DOPA Sparing Effect:
Once compound 40, a high mGlu4 selective and brain penetrant mGlu4 PAM, was found, its PD treating ability was investigated. An experiment was conducted to test out the PD treating ability of compound 40. 6-OHDA PD modeled rats were treated by either L-DOPA alone or in combination with compound 40. Results showed that L-DOPA alone successfully restored motor activity in rats back to normal levels at an optimal dosage of 20mg/kg. This L-DOPA dosage of 20mg/kg was found to increase rearing activity, a sign for potential LID development. The rats were then co-treated with a subthreshold (ineffective) L-DOPA dosage (6mg/kg) and compound 40, and results showed that motor activity was restored to the same level as the L-DOPA alone treatment at optimal dosage [Fig 2a]. This suggest that co-treatment with compound 40 generate a L-DOPA dosage sparing effect by decreasing 70% of the L-DOPA dosage required to fully restored motor activity in PD rat models. As mentioned previously, reducing L-DOPA dosage usage would reduce the occurrence of developing LID. This experiment successfully evidenced L-dopa sparing effect of compound 40, which could be used to reduce PD symptoms by expanding the upper limit of a safe and effective L-DOPA dosage 9.
dissertations.com/108/0500817.003.png" alt="" />
[Figure 2a: Total spontaneous locomotor activity measurements for 6-OHDA PD modeled mice treated by (mg/kg dosage) combinations of L-DOPA and Compound 40 in the following ratio (in (mg/kg)): [6:0], [6:1], [6:10], [6:30], and [20:0]] 9
Optimizing compound 40 to compound 60:
Once successfully demonstrated the high efficacy of compound 40 co-treatment in treating PD, compound 40 was further optimized to obtain a better brain penetration ability. Through a series of modifications, compound 40 was optimized into compound 60 (PXT002331), which could better penetrate the brain after oral dosage and generate similar mGlu4 PAM activity as compound 40. Compound 60 was also found to increase glutamate affinity for mGluR4 and had no effect on group I or group II mGluRs, suggesting its high selectivity against mGlu4 only 9.
A mGlu4 PAM in alleviate PD in primates
Not long after compound 60 (PXT002331) was found and tested in mice model, another paper came out that investigate the effect of compound 60 in primate PD models. In the past decade, there have been many studies that successfully showed the co-treatment of L-DOPA with a mGlu4 PAM could slow down the LID development, but this co-treatment effect hasn’t yet been tested in primates until recently. In a recent study, a group of researchers conducted experiments using the co-treatment of L-DOPA with mGlu4 PAM on primates. The purpose of testing this drug on primates is to ensure that the drug is brain-penetrant, safe for human, and is mGluR4 target specific with high efficacy in MPTP-lesioned macaque PD model 7.
Compound 60 Brain Penetrance Effect:
Brain penetrability of PXT002331 was tested by competing a radioligand (PXT012253) that binds the same mGlu4 allosteric site as PXT002331. PET scanning was done in macaque to track the radioligand affinity and results showed a reduction in radioactive ligand binding on mGlu4 after PXT002331 was introduced. This again confirm the brain penetrance ability and the high mGlu4 allosteric site affinity of PXT002331 after oral dosage 7.
Compound 60 prevent LID when co-treated with L-DOPA:
Similar to mouse study results, L-DOPA + PXT002331 co-treatment in primate PD model also restores the motor activity in macaque PD model. In this test, the macaque PD model used is characterized by a reduced locomotor activity that could be reversed upon L-DOPA treatment.
A co-treatment of L-DOPA with PXT002331 at an optimal oral dosage was found to increase the locomotor activity. And when only PXT002331 was treated, no effect on locomotor activity was observed. The fact that co-treatment restored the locomotor activity again reinforce the L-DOPA sparing characteristics of mGlu4 PAM PXT0023317.
Compound 60 prolongs L-DOPA efficacy when co-treated with L-DOPA:
Another test was conducted to test the efficacy of the compound 60-L-DOPA co-treatment. PD symptoms was fully reversed at an optimal DOPA dosage. If lowers this dosage to a suboptimal L-DOPA dosage, PD symptoms was treated with same efficacy, but the effect only lasted half the time as optimal dosage. Interestingly, when compound 60 was co-treated with this suboptimal L-DOPA dosage, the anti-parkinsonian effect of L-DOPA was found to be longer, suggesting that L-DOPA co-treatment with compound 60 prolongs the L-DOPA effect while maintaining the same level of drug efficacy, which means that a lower L-DOPA dosage could be used to treat PD. Overall, this paper tested out the high brain penetrance of compound 60 and its L-DOPA-sparing and prolongation effect when treated with L-DOPA 7.
Why PAM and not orthosteric agonist
PAM vs Orthosteric Agonist
Up til this point, many might have wondered why researchers decided to target on mGluR4 via PAM and not orthosteric agonist that binds on the mGluR4 active site. To answer this question, a group of researchers tested out the two types of mGluR4 agonists on 6-OHDA PD modeled rats. VU0364770 was used as the mGluR4 PAM and Lsp1-2111 was used as mGluR4 orchestic agonist. The very first test performed was to test if either has anti-dyskinetic activity, and not surprisingly, neither alone could reduce LID AIMs incidents or severity6. The next step was to test the two ligands’ effects on PD mice when co-treated with L-DOPA.
In the co-treatment experiment, 6-OHDA lesioned rats were first treated by L-DOPA. mGluR4 PAM VU0364770 and orthosteric agonist Lsp1-2111 were then co-treated with full L-DOPA dosage on mice with dyskinesia, as these drugs alone don’t produce anti-dyskinetic effect. The AIMs were then scored to see the effects of both co-treatment. Results show that L-DOPA co-treatment with either drugs did not produce an anti-dyskinetic effect 6. Many might be surprised why the mGluR4 and L-DOPA co-treatment didn’t produce anti-dyskinetic effect after. Well, this is because this particular experiment used a full L-DOPA dosage, instead of an acute, suboptimal-DOPA dosage. This again reinforces the idea that mGluR4 PAM co-treatment (with L-DOPA) works to treat PD by lowering the required L-DOPA dosage for full efficacy.
Nevertheless, another experiment was conducted that treats 6-OHDA lesioned rats with either the mGluR4 PAM or orthosteric agonist with L-DOPA at a subthreshold dosage. The L-DOPA subthreshold dosage means that L-DOPA treatment alone is ineffective. Then, a series of behavioral tests were performed on the rats. Specifically, rotational and rotarod behavioral tests were performed 6.
Both L-DOPA and PAM alone did not lead to significant increase of rotational activity, but co-treatment of PAM with L-DOPA increased rotational activity. L-DOPA alone at subthreshold dosage slightly improved rotarod performance, but its co-treatment with the PAM significantly increased rotarod performance. This suggest that a smaller L-DOPA dosage would be required (to produce the same degree of effect as a higher L-DOPA dosage) if it’s treated with the mGlu4 PAM VU0364770. In the rotational test for the LSP1-2111 orthosteric agonist, both L-DOPA and LSP1-2111 alone or their co-treatment did not lead to significant increase of rotational activity. As for the rotarod test, subthreshold dosage L-DOPA alone increase rotarod activity, but its co-treatment with LSP1-2111 did not produce additional increase of rotarod performance. These results suggest that the PAM VU0364770 produce a DOPA-sparing effect, whereas the LSP1-2111 lack this ability. This DOPA-sparing effect of VU0364770 was possibly due to its higher selectivity for mGlu4 (as to other group III mGluRs) than the LSP1-2111. For example, treating LPSP-2111 to the circulatory system could activate mGlu7 in the SNr, which lead to impacts that are bad for anti-dyskinetic effects. Whereas the mGlu4 PAM VU0364770 was tested to be highly selective for mGlu4 only 6.
Overall, both compounds were tested to not produce anti-LID effect. But unlike LSP1-2111, VU0364770 had a L-DOPA-sparing effect, meaning co-treatment of L-DOPA with VU0364770 reduces the L-DOPA’s dosage requirement for anti-dyskinesia effect. And since prevalence of LID depend on L-DOPA dosage, reducing the dosage of L-DOPA could reduce LID incidents 6.
Because of their ability in mediating the GABA and glutamate transmissions (found to be hyperactive in PD) in the basal ganglia, mGluRs have been a potential candidate to target for treating PD 1.
In PD and LID symptoms, overactive glutamate synaptic transmission was observed. Therefore, two main mGluR targets were heavily investigated on and were proven to successfully generate anti-dyskinetic effects: (1) inhibiting mGluR5 with antagonists to decrease the signaling downstream of excessive glutamate by reducing the nigral GABA level AND (2) activating mGluR4 with PAMs to decrease glutamate release in the sNr. More specially, it is the mGluR4 PAMs co-treatment that lead to anti-dyskinetic effect2,8. mGluR4 PAM and L-DOPA co-treatment was found to have two main benefits: (1) provide a L-DOPA sparing effect, which lowers LID incidents, AND (2) prolong L-DOPA efficacy 7. One main reason why mGluR4 was targeted by PAM and not orthosteric agonist is because only the PAM version of mGluR4 agonist could provide a L-DOPA sparing effect that is crucial to treating PD 6.
When comparing mGluR5 and mGLuR4 treatment to treat PD, mGluR4 is probably a better way to lower the excessive glutamate in PD symptoms for two reasons. First, it is more effective to directly decrease glutamate release by activating the presynaptic receptor than to inhibit the postsynaptic glutamate releasing receptors downstream. Second, mGluR4 PAMs has better potency than antagonists and NEM because mGluR4 PAMs was found to only need to occupy less than 50% of the brain mGluR4 to be effective, whereas antagonists and NEMs need to occupy more than 65% of their target mGluRs to be fully effective. Moreover, mGluR4 PAM PXT002331 has already been approved to enter phase IIa clinical trials. It is the first compound of its trial to enter phase IIa clinical trials on human. Much of the mystery is about to unravel as this mGluR4 PAM’s ability in treating PD symptoms is being tested in human 9.
- “Statistics.” Parkinson’s Foundation, 5 Dec. 2018, parkinson.org/Understanding-Parkinsons/Causes-and-Statistics/Statistics.
- Duty S. 2010. Therapeutic potential of targeting group III metabotropic glutamate receptors in the treatment of Parkinson’s disease. Br J Pharmacol. 161:271–287.
- Hopkins CR, Lindsley CW, Niswender CM., et al. mGluR4-positive allosteric modulation as potential treatment for Parkinson’s disease. Future Med Chem. 2009;1(3):501–513
- “Metabotropic Glutamate Receptor.” Wikipedia, Wikimedia Foundation, 9 Dec. 2018, en.wikipedia.org/wiki/Metabotropic_glutamate_receptor#cite_note-Endoh-9.
- Claudia Volpi, Francesca Fallarino, Giada Mondanelli, Antonio Macchiarulo & Ursula Grohmann et al. (2018) Opportunities and challenges in drug discovery targeting metabotropic glutamate receptor 4, Expert Opinion on Drug Discovery, 13:5, 411-423, DOI: 10.1080/17460441.2018.1443076
- Iderberg H, Maslava N, Thompson AD, Bubser M, Niswender CM, Hopkins CR, et al. Pharmacological stimulation of metabotropic glutamate receptor type 4 in a rat model of Parkinson’s disease and L-DOPA-induced dyskinesia: Comparison between a positive allosteric modulator and an orthosteric agonist. Neuropharmacology. 2015;95:121–9. doi: 10.1016/j.neuropharm.2015.02.023.
- Charvin, Delphine, et al. “An mGlu4‐Positive Allosteric Modulator Alleviates Parkinsonism in Primates.” The Canadian Journal of Chemical Engineering, Wiley-Blackwell, 14 Sept. 2018, onlinelibrary.wiley.com/doi/full/10.1002/mds.27462.
- Mela, Flora, et al. “Antagonism of Metabotropic Glutamate Receptor Type 5 Attenuates l‐DOPA‐Induced Dyskinesia and Its Molecular and Neurochemical Correlates in a Rat Model of Parkinson’s Disease.” The Canadian Journal of Chemical Engineering, Wiley-Blackwell, 22 Jan. 2007, onlinelibrary.wiley.com/doi/full/10.1111/j.1471-4159.2007.04456.x. Accessed 12 Dec. 2018.
- “Discovery, Structure–Activity Relationship, and Antiparkinsonian Effect of a Potent and Brain-Penetrant Chemical Series of Positive Allosteric Modulators of Metabotropic Glutamate Receptor 4.” ACS Publications, pubs.acs.org/doi/10.1021/acs.jmedchem.7b00991.
- Valenti, Ornella, et al. “Group III Metabotropic Glutamate Receptor-Mediated Modulation of the Striatopallidal Synapse.” Journal of Neuroscience, Society for Neuroscience, 6 Aug. 2003, www.jneurosci.org/content/23/18/7218?ijkey=01de60d8903c56f018830dca2409ea36285ccabb&keytype2=tf_ipsecsha.